This map shows predictions of habitat suitability for several biological groups of deep-sea corals.
These data are modeled predictions of potential habitat suitability based on a limited number of observations of coral presence and on predictive environmental characteristics. Therefore, the data reflect any sampling biases associated with a limited number of observations of coral presence in the Northeast region. The National Oceanic and Atmospheric Administration (NOAA) produced these data products using Maximum Entropy predictive habitat modeling (MaxEnt), which is a machine-learning technique widely used for presence-only data. The relationships between coral presence and habitat variables such as depth, bottom temperature, bottom salinity, and sediment grain size were modeled to produce regional maps of predicted habitat suitability. Model predictions are organized into five hierarchical categories of calibrated suitability thresholds: low, medium-low, medium, high, and very high. These calibrated thresholds enable comparisons or combinations of maps across coral taxa.
These data are hosted by NOAA Coastal Services Center and are complete.